temporal series of evi/modis to identify land converted to sugarcane

Revista Brasileira de Geografia Física 03 (2011) 562-574

Rocha, D. R. S.; Barbosa, H. A.; Silva, L. R. M. 562

ISSN:1984-2295

Revista Brasileira de

Geografia Física

Homepage: www.ufpe.br/rbgfe

A Gis Approach Using Remote Sensing Derived Products for Quantification of

Sugar Cane Productivity in Brazil

Diego Raoni da Silva Rocha1, Humberto Alves Barbosa

2, Leandro Rodrigo Macedo da Silva

3

1Universidade Federal de Alagoas (UFAL), Laboratory of Analysis and Processing of Satellite Images, Campus A. C.

Simões - Av. Lourival Melo Mota, s/n, Tabuleiro do Martins - Maceió - AL, CEP: 57072-970. E-mail:

[email protected]. 2Universidade Federal de Alagoas (UFAL), Laboratory of Analysis and Processing of Satellite Images, Campus A. C.

Simões - Av. Lourival Melo Mota, s/n, Tabuleiro do Martins - Maceió - AL, CEP: 57072-970. e-

mail:[email protected]. 3Universidade Federal de Alagoas (UFAL), Laboratory of Analysis and Processing of Satellite Images, Campus A. C. Simões - Av. Lourival Melo Mota, s/n, Tabuleiro do Martins - Maceió - AL, CEP: 57072-970. e-mail:

[email protected].

Artigo recebido em 15/08/2011 e aceite em 10/09/2011

R E S U M O

Este estudo teve como objetivo de desenvolver uma metodologia baseada na aplicação de produtos GEONETCast-

EUMETCast para estimativa da produtividade da cana-de-açúcar utilizando-se de um modelo agrometeorológico-

espectral. O estudo foi desenvolvido no município de Coruripe, localizado no estado de Alagoas, Brasil. O teste foi

realizado num período de cinco meses, abril a agosto, do ano de 2010. Conclui-se que a metodologia utilizada indica ser

útil para o apoio operacional de estimativa da produtividade da cana-de-açúcar, fornecendo valores médios de 37 a 40

t/ha.

Palavras-chave: Spot Vegetation, Meteosat-9, Produtividade Safra, Ilwis

A B S T R A C T

This study aimed to develop a GEONETCast-EUMETCast product-based method of estimating the productivity of cane

sugar using an agrometeorological-spectral model. The study was carried out in the Municipality of Coruripe, located in

the state of Alagoas, Brazil. The test was performed over the period of five months, from April to August of 2010. It

was concluded that the methodology is useful for developing estimates of operational support for the cane sugar

productivity, providing mean values of 37 to 40 t/ha.

Keywords: Spot Vegetation, Meteosat-9, Crop Yield, Ilwis

1. Introduction

1.1 The gap between science and practical

agricultural management

Current management practices in

Brazil in the field of agricultural management

are very often still based on outdated

knowledge and technology. Similarly to what

happens in many other regions of the world,

frequently scientists plan and develop their

methods in isolation, not grasping what is

really required by relevant stakeholders. On

the other hand, stakeholders are often

unaware of what science or knowledge-based

alternatives are available. Scientific research

is further isolated by lack of proven utility

*E-mail: [email protected] (Rocha, D. R. S.).



and inadequate representation of results,

whilst agricultural policy and management is

isolated by legal and professional precedence.

1.2 Making crop modeling useful for

decision-making: what output is needed, and

what input data are required to achieve the

modelling goals

In the scientific community frequently

model performance is evaluated using

procedures and statistical indicators which do

not necessarily reflect the usefulness for

practical decision-making in the region of

interest. Improved awareness of stakeholders’

real needs can help scientists to better orient

their work. On the other hand, typical current

performance rates of crop modelling

applications in the region may not be

sufficient for practical decisionmaking.

Low model performance is often a

consequence of limited input data sets on

which the model application is based.

However, over the past decades an increasing

amount of potentially useful products derived

from remote sensing have become available,

but their potential for improving model

performance in the region has not been well

assessed yet.

1.3 Agro-meteorological parameters from

satellite remote sensing products and GIS

approach

Remote sensing images play an

important role in agricultural crop production

over large area, quantitatively and

nondestructively, because agricultural crops

are often difficult to access, and the cost of

ground based estimation of productivity can

be high. The recent development of

GEONETCast–EUMETCast data has

provided the capability to obtain frequent and

accurate measurements of a number of basic

agro-meteorological parameters (e.g.

evapotranspiration, surface albedo, surface

temperature, solar radiation, rainfall etc.). The

GEONETCast real-time data dissemination

system represents a global network of

communication satellite based data

dissemination system to distribute space-

based, air-borne and in situ data, metadata and

products to diverse user communities. The

satellite estimated agro-meteorological

parameters provide a complete and spatially

dense observations capability for assessing

regional vegetation potential productivity

(Barbosa et al., 2006; Barbosa et al., 2009).

With a Geographic Information System (GIS)

approach, it provides a framework to ingest in

the analysis the information of

agrometeorological parameters that influence

crop yield.

1.4 Sugar cane crops in Brazil

Agricultural production in the semi-

arid region of Brazil is characterized by

extensive, low investment subsistence

farming focused on limiting the impacts of

hydrological and policy risks. As a

consequence, agricultural, labor and natural

resource productivity has remained low,

however at high environmental costs,



especially with respect to land degradation

and loss of natural resources and biodiversity.

Agriculture is the primary social and

productive economic sector in the semi-arid

Latin American countries and forms the basis

for rural welfare and food security, and the

platform for structural change and economic

take off towards sustainable socio-economic

development and growth. Sugar cane is one of

the most important cash crops in Brazil. It is

an annual crop with solid jointed stems and its

mode of photosynthesis is very efficient and

growth is quick. Brazil is the biggest grower

of sugarcane, which is used to produce sugar

and ethanol gasoline-ethanol blends gasohol

for transportation fuel. Its bio fuel industry is

a leader, a policy model for other countries,

and sugarcane ethanol is considered the most

successful alternative fuel to date.

1.5 General objective

The general objective of this

application is to increase the role of crop

modelling applications for estimating sugar

cane productivity in Brazil by incorporating

satellite remote sensing products to evaluate

estimation of productivity of sugar cane crops.

A number of tailored and GIS-compatible

products are generated and integrated to

quantify sugar cane productivity.

1.6 Specific objective

For this application products that are

disseminated through GEONETCast (NDVI

S–10, SPOT Vegetation DMP and ETp) are

used to develop a remote sensing based

approach to improve sugar cane productivity

estimation for the municipality of Coruripe,

located in Alagoas, Brazil. The study is

conducted for the period of April to August,

2010. The sugar cane productivity is derived

in 9 computational steps.

2. Methodology

2.1 Data Used

2.1.1 Local/Regional (in-situ) data

The application focuses on the sugar cane

plantation for the entire Coruripe

municipality, located in Alagoas State, Brazil.

The sugarcane parameters used are:

BF = Respiration Factor (0.5 for

temperatures ≥ 20°C and 0.6 for

temperatures <20°C), after Gouvêa

(2008);

APF = Agricultural Productivity

Factor (2.9), after Ruddorf (1985);

Ky= yield response factor, after

Doorenbos et al (1979);

Kc= Crop coefficient.

The crop coefficient is defined as the

ratio of crop evapotranspiration (Etc) with

respect to reference evapotranspiration (ET0).

Kc is crop specific and is depending on the

crop growth stage. Crop evapotranspiration at

any time during the growing season is the

product of reference evapotranspiration and

the crop coefficient, expressed as: ETc = ET0

* Kc. Crop coefficients have been developed

for nearly all crops by measuring crop water

use with lysimeters and dividing the crop



water use by reference evapotranspiration for

each day during the growing season.

2.2 Products used from GEONETCast

The following data delivered through

GEONETCast has been used for the period

April to August, 2010:

Remote sensing images: SPOT

Vegetation–2 and SPOT–5, SEVIRI

instrument on METEOSAT–9;

Remote sensing based products:

o NDVI S10 (NDVI) and Dry

Matter Productivity (DMP)

having a spatial resolution of

1Km.;

o LSA–SAF ETp product for

South America.

Source of the data is the

GEONETCast ground receiving station

installed at the Laboratory of Analysis and

Processing of Satellite Images (LAPIS),

Federal University of Alagoas (UFAL) (see:

http://www.lapismet.com) and the SPOT

Vegetation products are obtained from the

archive, maintained by VITO (see:

http://free.vgt.vito.be/).

This application proposes to test a

remote sensing approach to quantify estimates

of sugar cane productivity over the Coruripe

municipality and sugar cane crop estimates

are computed for each map pixel using NDVI,

DMP and Etp images by applying radiative,

aerodynamic and energy balance physics

using 9 computational steps. The flow

diagram of the methodology to quantify sugar

cane productivity using these satellite derived

products is shown in figure 1.

2.3 Data pre-processing for quantification

sugar cane productivity

Ensure that you have unzipped the

exercise data and move using the ILWIS

navigator to this active working directory.

Close ILWIS and open ILWIS again and

ensure that the path to your active working

directory is correct. In order to minimize the

data load for this exercise all time series data

(NDVI, with and without status mask, DMP,

ETp_avg and ETp_std) have been pre-

processed and sub maps for the Coruripe

Municipality have been created. The pre-

processing steps are described in the

following section.

2.4 Step1: Input NDVI and DMP databases

using algorithm adapted from GEONETCast

Toolbox

To implement a methodology to

import both the NDVI and DMP time series

raster data into ILWIS, specific routines

available within the GEONETCast–Toobox

are adapted for import. Upon import also the

status mask was applied, to retain only the

map values that meet the flag criteria, such as

cloud free, land pixels, good radiometric

quality in the Red and NIR bands. For further

details see also Chapter 4.5.1.2 as well as

Maathuis et al, 2011.



Figure 1: Flow chart of the methodology adapted

Open the MapList

“NDVI_Coruripe_Apr_Aug” and display the

time series as an animated map sequence,

using as Representation “NDVI1”. Note that

there are 15 maps and each map is

representing the decadal NDVI computed

using a Maximum Value Composite

Algorithm. Map layer 1 represents the 1st

decade of April 2010 and map layer 15 is the

last decade of August 2010. Move the mouse,

keeping the right mouse button pressed, over

the active map display window and note the

values. From the menu of the active map

display window, select the option “Layers =>

Add Layer” and select the Polygon Map

“coruripe”, in the Display Options window

unselect “Info” and select the option

“Boundary Only” and press “OK” to see the

municipal boundary. Close the Active map

window and now display the time series

“NDVI_SM” in a similar manner. Here the

status flags have been applied and those

pixels that did not meet the flag criteria have

been re-assigned as “no-data”, represented by

a “?”.

Step 2: Computation of FVC from NDVI

The FVC is the one biophysical

parameter that determines the contribution

partitioning between bare soil and vegetation

for surface evapotranspiration,

photosynthesis, albedo, and other fluxes

crucial to land–atmosphere interactions. The

NDVI needs to be converted to FVC. Here



use is made of the formula proposed by

Muñoz et al (2005), see equation 1. The flag

corrected NDVI time series is going to be

used.

FVC = 1.1101 * NDVI - 0.0857 (Eq 1)

To carry out this calculation for each

NDVI map within the time series open from

the main ILWIS menu the option “Operations

=> Raster Operations => MapList

Calculation” and provide the information as

given in figure 6.2.

Press “Show” to execute the operation,

display the resulting time series as an

animated sequence, use as Representation

“fvc” and check the values obtained using the

mouse, keeping the right mouse button

pressed, over the active map display window.

Step 3: Computation of LAI from FVC

The LAI, defined, as the total one-

sided leaf area per unit ground area, is one of

the most important parameters characterizing

a canopy. Because LAI most directly

quantifies the plant canopy structure, it is

highly related to a variety of canopy

processes, such as evapotranspiration,

interception, photosynthesis and respiration.

The FVC is converted to Leaf Area Index

(LAI) by means of the formula proposed by

Norman et al (2003), see equation 2.

LAI = - 2Ln (1 – FVC) (Eq 2)

To carry out this calculation for each FVC

map within the time series open from the

main ILWIS menu the option “Operations =>

Raster Operations => MapList Calculation”

and provide the information as given in figure

6.3.

Press “Show” to execute the operation,

display the resulting time series as an


“lai” and check the values obtained using the

mouse, keeping the right mouse button


Now also check the command line history

from the main ILWIS menu by pressing the

“drop down” button on the right hand side of

the command line. Note the command line

string that has been used to create the LAI

time series. It is given by the following string:

LAI.mpl = maplistcalculate ("-2*ln(1-

@1)",0,14,FVC.mpl)

Check the expression, compare it with figure

6.2. Note that in the following sections the

command line expressions to calculate a

maplist will be given.

Step 4: Computation of growth factor from

LAI

Experimental evidence indicated that

the growth rate of several agricultural crop

species increases linearly with increasing

amounts of LAI, when soil water nutrients are

not limiting (Doorenbos and Kassam, 1979).

Berka et al (2003) developed a simple

approach for deriving the growth rate from

the LAI, see equation 3.

CGF = 0,515 – e ( - 0, 667 – (0,515 * LAI))

(Eq 3)

Where: CGF = Corrected Growth Factor and

LAI = Leaf Area Index.

To derive the Corrected Growth Factor (CGF)

https://www.researchgate.net/publication/26365682_Soybean_yield_estimation_by_an_agrometeorological_model_in_a_GIS?el=1_x_8&enrichId=rgreq-c68b9135-19d2-4ea0-9f90-025d981ba74e&enrichSource=Y292ZXJQYWdlOzIyNDExNjc0NjtBUzo5OTc5ODQzODkwNzkwNkAxNDAwODA1MjEwMDUz



type the following equation on the ILWIS

command line:

CGF.mpl:=maplistcalculate("0.515-exp(-

0.667-(0.515*@1))",0,14,LAI.mpl)

Press “Enter” to execute the operation,

display the resulting CGF time series as an


“Pseudo” and check the values obtained using

the mouse, keeping the right mouse button


Step 5: Computation of maximum yield

potential (Yp)

The final equation that can be used to

derive maximum yield potential (Yp) is based

on an equation which includes evaporative

fraction corrected growth factor (CGF),

respiration factor (BF), agricultural

productivity factor (APF) and production of

dry matter (DMP) product.

Yp = CGF * BF * APF * DMP (Eq 4)

Where Yp is the maximum yield

potential (kg ha-1), BF is the respiration

factor (0.5 for temp. ≥ 20°C and 0.6 for temp

<20°C), APF is the agricultural productivity

factor (2.9) as described in Rudorff (1985)

and DMP is the Dry Matter Productivity

derived from Spot-Vegetation data. First

display using an animated sequence the time

series “DMP_04_08_coruripe” using as

Representation “Pseudo” and check the map

values using the mouse, keeping the right

mouse button pressed, over the active map

display window. Further information on this

product can be found in Bartholomé (2006),

the unit is kg/dry matter/ha/day. After you

have seen the animated sequence, close the

animation.

To calculate the maximum yield potential for

each time step, type the following expression

on the command line in the main ILWIS

menu:

Yp.mpl:=maplistcalculate("@1*0.5*10*2.9*

@2",0,14,CGF.mpl,DMP_04_08_Coruripe.m

pl)


display the resulting CGF time series as an


“Pseudo” and check the values obtained using

the mouse, keeping the right mouse button


Note that the factor 10 in the expression

above is used to convert from kg/dry

matter/ha/day to kg/dry matter/ha/decade!

Step 6: Retrieval of evapotranspiration (ETp)

via LSA –SAF ETp product

The crop coefficient is defined as the

ratio of crop evapotranspiration, ETr, to

reference evapotranspiration, ETp. Kc is crop

specific depending on the crop growth stage

and details are given in table 6.1. Crop

evapotranspiration at any time during the

growing season is the product of reference

evapotranspiration and the crop coefficient as

given by equation 5.

ETr = ETp * Kc (Eq 5)

By use of the newly developed LSA–

SAF products, one is now able to obtain

frequent and accurate measurements of a



number of basic agro-meteorological

parameters (e.g. surface albedo, surface

temperature, evapotranspiration). The satellite

estimated agro-meteorological parameters

have several advantages compared to

conventional measurements of

agrometeorological data using a scattered

ground meteorological observation network.

Open the map list “ETp_avg” and

display the time series as an animated

sequence, using as Representation “Pseudo”.

This maplist has been compiled processing

each 30 minutes LSA_SAF ET-product from

1 April to 31 August 2010. All products have

been added on a daily basis, corrected for the

time interval, as the product unit is expressed

in mm/hr, but the time step is half hourly

(therefore the daily sum has been divided by

2). For each decade the respective daily

products have been summed and the average

was computed to obtain the average ET per

decade. Also the standard deviation has been

computed and was aggregated to obtain the

decadal standard deviation. This map list is

called “ET_std. Display also this map list

using as Representation “Pseudo”.

To obtain the crop evapotranspiration

the following procedure has been adopted

using the time series ETp_avg and ET_std

data: For the months of April, May and June a

crop coefficient was used of 1.68:

ETr = (ETp_avg – ET_std) * 1.68 (Eq. 6)

For the months of July and August a crop

coefficient was used of 1.18:

ETr = (ETp_avg – ET_std) * 1.18 (Eq.

7)Type the following equations in the

command line of the main ILWIS menu and

press enter and “OK” to execute the

operations:

ed. From the main ILWIS menu, select “File

=> Create => MapList” and add all newly

created ETr* maps in a sequential order to the

right hand listing, using the “>” icon. Specify

as Map List name “ETr” and press “OK”.

Display the newly created map list “ETr” as

an animated sequence, using as

Representation “Pseudo”, check the map

values obtained.

Step 7: Estimation of sugar cane productivity

The sugarcane yield estimation model

over the growing season, on a decadal basis,

is accomplished by using an

agrometeorological model (Equation 8)

according to Doorenbos and Kassam (1979):

(Eq 8)

where Ye is the estimated yield (kg ha-1), Yp

the maximum yield (kg ha-1), ky the yield

response factor described in (Doorenbos and

Kassam, 1979); ETr the actual

evapotranspiration (mm) and ETp the

maximum evapotranspiration (mm).

Maximum yield (Yp) is established by the

genetic characteristics of the crop and by the

degree of crop adaptation to the environment

(Doorenbos and Kassam, 1979). The ky

factors used here are for April = 1.2, for May

= 1.3, for June = 1.2, for July = 1.1 and for



August = 1.1. To calculate the estimated yield

the following equations have to be entered

from the command line on the main ILWIS

menu:

Ye1:=Yp_1*(1-1.2*(1-

(ETr1/etp_avg_0401_coruripe)))

In the Raster Map Definition window, set the

minimum value range to “0” and the

“Precision” to “0.01” and press “OK” to

execute the operation. Display the map

calculated and check the values obtained.

Note that each pixel represents and area of 1

km2 and the estimated yield is expressed in

kg ha-1! Repeat the Ye calculations for the

other decades using the following set of

equations (note the change of the Yp) and

keep on setting the minimum map value to

“0” and use as Precision “0.01”:

After all calculations are performed a

new map list has to be created. From the main

ILWIS menu, select “File => Create =>

MapList” and add all newly created Ye* maps

in a sequential order to the right hand listing,

using the “>” icon. Specify as Map List name

“Ye” and press “OK”. Display the newly

created map list “Ye” as an animated

sequence, using as Representation “Pseudo”,

check the map values obtained. Close all

active map windows.

Step 8: Local mask of estimated yield

The resulting map list of the estimated

yield (Ye) should be clipped to the mask of

the Coruripe municipality boundaries in the

State of Alagoas, Brazil. Right click with the

mouse button on the polygon map “coruripe”

and from the context sensitive menu, select

the option ”Polygon to Raster”, as

GeoReference select

“CFG_Coruripe_Apr_Aug_1”, leave the

Output Raster Map name as indicated by

default and press “Show”.

Display the map and note the content.

Close the map, right click with the mouse on

the RasterMap “coruripe” andfrom the

context sensitive menu, select the option

“Properties”. Note that the domain here is

specified as Identifier “coruripe” and although

when looking at the map content the “1”

which was indicated is representing an

identifier number but not a value. In order to

use this map in a calculation we have to

change it to a value. Type the following

expression in the command line of the main

ILWIS menu:

mask:=iff(coruripe="1",1,0)

Note that from the Raster Map

Definition window the domain given is now a

“value” domain. Press “OK” to execute the

operation. Display the map “Mask” and note

the content. Now we can use the mask to

extract the “Ye” for the Coruripe

Municipality. Type the following expression

on the command line in the main ILWIS

menu:

Ye_Coruripe.mpl:=maplistcalculate("i

ff(mask=1,@1,?)",0,14,Ye.mpl)


display the resulting “Ye_Coruripe” time

series as an animated sequence, use as



Representation “Pseudo” and check the values

obtained using the mouse, keeping the right

mouse button pressed, over the active map

display window.

Step 9: Total Yield Productivity using a sugar

cane crop mask

The time series “Ye_Coruripe”

corresponds to the estimated sugar cane yield

on a decadal basis. To estimate Total Yield

Productivity for whole time series, each

decadal period from the Map List

Ye_Coruripe should be accumulated. Enter

the following expression on the command line

to obtain the sum of the time series and press

enter to execute the operation:

Ye_sum:=MapMaplistStatistics(Ye_C

oruripe.mpl, Sum, 0, 14)

Finally to estimate total yield productivity the

average amount of water (76 %) withtout

stress is added to the sugar cane and the initial

weight of the sugar cane stems during

planting has to be added as well (here a value

is used of 15 ton/ha). Enter the following

expression on the command line to obtain the

Total Yield Productivity and press enter to

execute the operation:

Ye_total:=Ye_sum*1.76+15000

Right click with the mouse button on the

polygon map “sugarcane_mask” and from the

context sensitive menu, select the option

”Polygon to Raster”, as GeoReference select

“CFG_Coruripe_Apr_Aug_1”, leave the

Output Raster Map name as indicated by

default and press “Show”. Display the map

and note the content.

Open from the main ILWIS menu the option

“Operations => Raster Operations => Cross”.

Select the raster map with the total yield

productivity, here called “Ye_total” as 1st

Map. Select raster map “sugarcane_mask” as

2nd Map. Type “yield_mask” as Output Table

Figure 2. Ye-total for the cururipe area and cross results using a sugar cane mask and press

“Show”. Note the content of the cross table. The figure below shows the final results of the analysis,

using the boundaries of the municipality and the sugar cane mask as overly on the Ye_total map.



3. Summary and Conclusions

For this exercise GEONETCast–

EUMETCast data (NDVI S-10, DMP SPOT

and ETo) is applied to test a remote sensing

approach to improve sugar cane Total Yield

Productivity over the Municipality of

Coruripe. The test is performed for the period

April to August, 2010 and the sugar cane

productivity is derived using nine

computational steps.

The results show that the methodology

adopted has three distinct advantages

compared to the generally accepted “kc x

ETo” method for computing ET. First

advantage is that the acreage of water-using

land is observed directly from the satellite

products, so accurate land use is implicit to

the process. Second, there is no need to

incorporate a crop type map to solve the

energy balance, so accurate records of

cropping patterns are not required.

These features overcome the typical

difficulty of assembling accurate records of

irrigated areas and cropping patterns,

especially for historical analyses. Thirdly, the

LSA-SAF ETo (aggregated) product can be

imported into a GIS for spatial analysis, either

alone or in combination with land use and

other spatial data, inherently accounting for

the effects of salinity, deficit irrigation or

water shortage, disease, poor plant stands and

other ET-reducing influences on the ET flux.

These influences are difficult to take into

consideration using the standard “kc x ETo”

computation. Furthermore the software tools

have demonstrated a great flexibility and ease

of use.

The results presented in this research

show the values of sugarcane production are

in the average range of 37 to 40 ton/ha at the

level of the municipality area during the test

period. The sugarcane area finally selected is

showing significantly higher values. This is a

first indication where the sugar cane areas can

be harvested. Results are very encouraging,

through a high spatial variability of crop yield

is found, suggesting adjustments are needed

to transform the original satellite data-based

scheme into satellite estimated agro-

meteorological parameters. Further studies are

required to analyse these results into more

detail as these depend for example on the

spatial resolution of the input background

fields, their physical content and many other

factors. Furthermore there is a need to use a

longer time series and analyse into more

detail the temporal response of e.g. the DMP

of the study area.

These finding present a first step

towards an operational use of ILWIS in Brazil

using NDVI S-10, DMP SPOT and ET0 for

operational estimating of sugar cane

productivity. This preliminary assessment

demonstrates the feasibility of the proposed

methodology to be useful in homogenous

areas with the same characteristics and to

focus on the control factors and incorporate

local information to enable better model

calibration and thus improve the results.



4. References

Allen, R. G.; Pereira, L. S.; Raes, D.; Smith,

M. (1998). Crop Evapotranspiration:

Guidelines for Computing Cropwater

Requirements. Rome: FAO. Irrigation end

Drainage Paper, 56. 300p.

Bartholomé, E. (editor) (2006). VGT4 Africa

user Manual. 1st Edition. EUR 22344 EN,

European Communities. Available online at:

http://www.devcocast.eu/ViewContent.do?pa

geId=40.

Berka, L.M.S.; Rudorff, B.F.T.;

Shimabukuro, Y.E. (2003). Soybean yield

estimation by an agrometeorological model in

a GIS. Scientia Agricola, v.60, p.433-440.

Barbosa, H.; Huete, A; Baethghen, W. (2006).

A 20-year study of NDVI variability over the

Northeast Region of Brazil. Journal of Arid

Environments ,67, 288-307.

Barbosa, H. A. (2009). Linking Sustainable

Indices and Climate Variability In The State

Of Ceará, Northeast Brazil. In: The Future of

Drylands, 6, 435-445, DOI: 10.1007/978-1-

4020-6970-3_40.

Choudhury, B.J.; Ahmed N.U.; Idso S.B.;

Reginato R.J.; Daughtry, C.S.T. (1994).

Relations Between Evaporation Coefficients

and Vegetation Indices Studied By Model

Simulations. Remote Sensing of

Environment, v.50, p.1-17.

Doorenbos, J.; Kassam, A.H. (1979): Yield

response to water. Food and agriculture

organization of the United Nations, fao –

irrigation and drainage paper no. 33, 193 p.

Gouvêa, J.R.F. (2008). Mudanças climáticas e

a expectativa de seus impactos da cultura de

cana-de-açúcar na região de Piracicaba – SP.

Dissertação (Mestrado em Física do Ambiente

Agrícola) – Escola Superior de Agricultura

“Luiz de Queiroz”, Universidade de São

Paulo, Piracicaba. 99 p.

Muñoz, J. C. J.; Sobrino, J.A. Guanter, L.;

Moreno, J.; Plaza, A.; Matínez, P. (2005).

Fractional Vegetation Cover Estimation from

Proba/CHRIS Data:Methods, Analysis of

Angular Effects and Application to the Land

Surface Emissivity Retrieval. Proc. 3rd ESA

CHRIS/PROBA Workshop, ESRIN, ESA SP-

593, Frascati, Italy, June 2005

Maathuis, B., Mannaerts, C.; Retsios, B.,

Schouwenburg, M. And Lemmens, R. (2011).

Geonetcast Toolbox. Installation,

configuration and user guide of the

GEONETCast Toolbox plug-in for ILWSI

3.7, XML version 1.2. ITC. Enschede, The

Netherlands. 2011. http://52north.org/down

loads/earth-observation/geonetcast/toolbox.

Norman, J.M., Anderson, M.C., Kustas, W.P.,

French, A.N., Mecikalski, J., Torn, R., Diak,

G.R., Achmugge, T.J. (2003). Remote

Sensing of Evapotranspiration for Precision-

Farming Applications. In: IEEE International

Geoscience and Remote Sensing Symposium,

http://52north.org/down




https://www.researchgate.net/publication/226782444_Linking_Sustainable_Indices_and_Climate_Variability_in_the_State_of_Ceara_Northeast_Brazil?el=1_x_8&enrichId=rgreq-c68b9135-19d2-4ea0-9f90-025d981ba74e&enrichSource=Y292ZXJQYWdlOzIyNDExNjc0NjtBUzo5OTc5ODQzODkwNzkwNkAxNDAwODA1MjEwMDUz







Toulouse. Proceedings IGARSS 2003 –

Organizing Committee, 2003. p. 21-25.

Rudorff, B.F.T. (1985). Dados Landsat na

estimativa da produtividade agrícola da cana-

de-açúcar. Dissertação (Mestrado em

Sensoriamento Remoto) – Instituto Nacional

de Pesquisas Espaciais - INPE, São José dos

Campos. 114 p.

Toledo Filho, M.R. (1988). Probabilidade de

suprimento da demanda hídrica ideal da

cultura de cana-de-açúcar (Saccharum spp.)

através da precipitação pluvial na zona

canavieira do estado de Alagoas. 1988.72f.

Dissertação (Mestrado) – Progama de Pós-

graduação em Agronomia, Escola Superior de

Agricultura “Luiz de Queiroz”, Universidade

de São Paulo, Piracicaba.

https://www.researchgate.net/publication/43129046_Dados_Landsat_na_estimativa_da_produtividade_agricola_da_cana-de-acucar?el=1_x_8&enrichId=rgreq-c68b9135-19d2-4ea0-9f90-025d981ba74e&enrichSource=Y292ZXJQYWdlOzIyNDExNjc0NjtBUzo5OTc5ODQzODkwNzkwNkAxNDAwODA1MjEwMDUz







temporal series of evi/modis to identify land converted to sugarcane

Documents